The nave antibody repertoire is often viewed as a random sampling of the enormous diversity of antigen binding sites created by V(D)J rearrangement and N addition, especially in the heavy chain CDR3 (CDR-H3) interval that lies at the center of the antigen binding site. Inconsistent with this presumption of randomness is the finding of clear biases in amino acid composition that influence the hydrophobicity and structure of CDR- H3. These observations have led us to the over-all hypothesis that categorical regulation of CDR-H3 amino acid content affects the likelihood of specific epitope recognition, which will influence responses to vaccines, self-antigens and pathogens. HIV is an example of a pathogen that may require broadly protective antibodies to undergo 30% to 40% somatic mutation, taking years to produce. Elucidation of the somatic mechanisms that regulate the diversity of the repertoire is likely to identify new pathways for its therapeutic manipulation. Dysregulation of these mechanisms may also contribute to autoimmunity and vaccine failure. In this application we focus on one of the earliest stages where the amino acid composition of CDR-H3 appears to be constrained; the pre-B cell receptor (pre-BCR) checkpoint. High resolution structural analysis of the pre-BCR has defined a sensing site where the Ig? HC interacts with CDR-H3. We will test the hypothesis that the pre- BCR uses the CDR-H3 sensing site to select against hydrophobic amino acids typically encoded by DH reading frame 2. We will use an existing panel of D-altered mice and a new panel of VpreB1 altered mice to determine how alteration of CDR-H3 or VpreB1 sequence affects the outcome of passage through the pre-BCR selection checkpoint. We will evaluate pre BCR expression in pre-B cells, and B cell development in the bone marrow and periphery of mutant mice and compare them to wild type. We will evaluate patterns of apoptosis and cell cycle, both of which should be altered if our hypothesis is correct. We will clone and analyze VDJC??transcripts from living and apoptotic pr B cells from the DH or VpreB1 altered mice and compare them to wild type. We predict that alteration of DH sequence to force use of hydrophobic amino acids typically found in reading frame 2 will result in increased apoptosis and diminished cell division. On the other hand, alteration of VpreB1 will promote increased survival of B cells using CDR-H3s enriched for hydrophobic amino acids. We will extend our studies to humans by evaluating the CDR-H3 repertoire in early and late pre-B cells from human bone marrow to test if this same mechanism of pre-BCR selection is operating to constrain the repertoire. The technical innovation is our use of gene-targeted mice to test the hypothesis that the pre-BCR uses the CDR-H3 sensing site to select against hydrophobic amino acids. We expect that this study will reveal the existence of a previously speculated but, until now, untested mechanism of selection by the pre-BCR. In future studies, mice generated in this project will be used to explore the role of pre-BCR selection in regulating epitope recognition and thus shaping the immune response to both self and non-self antigens.